Approximately 1 in 800 live born babies are found to have Down syndrome (complete or partial trisomy 21) [1]. The risk of delivering a baby with Down syndrome (DS) increases with increasing maternal age [2]. By offering amniocentesis to women 35 and over, approximately 20% of DS pregnancies can be detected. It has been observed that low maternal serum alpha fetoprotein (MSAFP) levels are associated with DS, the risk increasing with decreasing MSAFP levels [3,4]. When maternal age and MSAFP level risks are combined, the rate of DS detection can be improved to 40% with some increase in efficiency. MSAFP screening is now used routinely to detect DS as well as for its original purpose to detect neural tube defects. Two other maternal serum biochemical markers have recently been reported to be altered in DS pregnancies. Human Beta chorionic gonadotropin (BetahCG) levels tend to be raised, [5-8] and unconjugated estriol (UE) levels tend to be lowered [9,10]. It is estimated that by combining risks associated with MSAFP, BetahCG and UE concentrations and maternal age, the detection rate could be raised to at least 60% and the efficiency improved [6]. The studies so far reported used look - back data and stored or post amniocentesis samples [5-12]. Several regional MSAFP screening programs in the Great Lakes Regional Genetics Group wish to do a combined prospective study in order to assess risks accurately before expanding their prenatal testing. It is proposed to obtain serum samples from patients who have already opted to undergo amniocentesis because of perceived increased risk of chromosome abnormalities. These would be collected at each of the participating centers and mailed to a single laboratory where MSAFP, BetahCG and UE levels would be assayed. Information regarding fetal karyotype and subsequent pregnancy out-come could be forwarded to the same center where this data would be linked to the biochemical data. The anonymous data would be statistically analyzed in an attempt to answer the following questions. 1. How independent is each of the four variables associated with increased DS risk? 2. What proportion of DS would have been detected by any combination of the variables and with what degree of efficiency? 3. Is there an optimum gestational age for detection? 4. What other abnormalities can be detected? 5. Can any problems associated with poor pregnancy outcome be predicted? Depending on the answer to 1 the study would be curtailed, expanded and/or prolonged. An alogarithm for calculating risk would be developed and a computer program written which would then be available to participating centers.